R Notebook

library(haven)
Data <- read.csv("/media/empro/5C02C2EB02C2C8EA/Ubuntu/Fendler/ovaryAnalysis/processed.csv", header = TRUE)
#View(Data)

Labels <- Data[c(1)]
Data <- Data[-c(1)]

zbadamy macierz korelacji

corMatrix <- cor(Data, method = 'spearman')

#heatm <- heatmap(corMatrix)
heatm <- heatmap(corMatrix, Rowv = NA, Colv = NA)

Mediany

library(data.table)

data.table 1.10.4.3
  The fastest way to learn (by data.table authors): https://www.datacamp.com/courses/data-analysis-the-data-table-way
  Documentation: ?data.table, example(data.table) and browseVignettes("data.table")
  Release notes, videos and slides: http://r-datatable.com

View(Labels)

Sprawdz dystrybucje

library(ggplot2)
for(colmn in Data)
{
  hist(colmn)
}

toM <- subset(Data, Labels[c(1)] == 'can')
#View(toM)
#toM <- setDT(Data, key=Labels)['can']

policz mediane

names <- c()
meds <- c()
ps <- c()
for(i in 1:ncol(toM))
{
  med = median(toM[, i])
  #print(median(as.numeric(as.vector(toM[c(i)]))) )
  name = colnames(toM)[i]
  p = wilcox.test(Data[, i] ~ Labels[, 1])$p.value
  
  #cat("Nazwa: ", name, "  mediana: ", med, "  p: ", p, "\n")
  
  names <- c(names, name)
  meds <- c(meds, med)
  ps <- c(ps, p)
}

#meds <- log2(meds)
volData = data.frame(names, meds, ps)
#View(head(volData, n=10) )
volData <- volData[order(volData$ps),]
View(head(volData, n=10) )

Beniamini-Hochberg Procedure - NOPE FDR

bh = p.adjust(volData$ps, method = 'fdr')
#View(head(bh, n=10))
volData <- data.frame(volData, bh)
View(head(volData, n=10))

Rysujemy piękny volcanoPlot

library(dplyr)

Attaching package: ‘dplyr’

The following objects are masked from ‘package:data.table’:

    between, first, last

The following objects are masked from ‘package:stats’:

    filter, lag

The following objects are masked from ‘package:base’:

    intersect, setdiff, setequal, union

library(ggplot2)

View(volData)

drawRes = mutate(volData, sig=ifelse(volData$bh < 0.05, 'FDR<0.05', 'Not Sig'))
volPlo = ggplot(drawRes, aes(meds, -log10(ps))) + geom_point(aes(col=sig)) + scale_color_manual(values = c('red', 'black'))
#0volPlo <- volPlo + geom_text(data=filter(drawRes, bh<0.05), aes(label=names))
volPlo <- volPlo + scale_x_continuous(limits = c(-2, 2))
volPlo

move labels

library(ggrepel)

volPlo <- volPlo + geom_text_repel(data=filter(drawRes, bh < 0.05), aes(label=names))
volPlo

Machine Learning